Ochotonidae, Lagomorpha) in the Pleistocene of Sardinia (Italy)

Home , Nuralagus, Prolagus, Rhagamys

Rivista Italiana di Paleontologia e Stratigrafia (Research in Paleontology and Stratigraphy) vol. 122(1): 25-36. March 2016

COMPARING THE BODY MASS VARIATIONS IN ENDEMIC INSULAR SPECIES OF THE GENUS PROLAGUS (OCHOTONIDAE, LAGOMORPHA) IN THE PLEISTOCENE OF SARDINIA (ITALY)

BLANCA MONCUNILL-SOLÉ1, CATERINELLA TUVERI2, MARISA ARCA2 & CHIARA ANGELONE1*

1Institut Català de Paleontologia Miquel Crusafont, Edifici Z ICTA−ICP, Carrer de les Columnes s/n, Campus de la Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain. E-mail: [email protected], [email protected]. *Corresponding author. 2Soprintendenza Archeologia della Sardegna, via G. Asproni 33, 08100 Nuoro, Italy. E-mail: [email protected], marisa.arca@ beniculturali.it.

To cite this article: Moncunill-Solé B., Tuveri C., Arca M. & Angelone C. (2016) - Comparing the body mass variations in endemic insular species of the genus Prolagus (Ochotonidae, Lagomorpha) in the Pleistocene of Sardinia (Italy). Riv. It. Paleont. Strat. 122(1): 25-36

Keywords: Body mass, regression models, postcranial bones, Island Rule, Mediterranean islands, Prolagus figaro, P. sardus.

Abstract. Prolagus figaro and P. sardus are part of an endemic insular anagenetic lineage that populated Sardinia since the earliest Late Pliocene to Holocene. BM of some populations of these two species was calculated using regression models. The best BM proxies for Prolagus are: femur length, zeugopod measurements and distal humerus diameter. The anagenetic lineage shows a BM increase of ca 20% from the populations of P. figaro (398-436 g) to P. sardus (504-525 g). The trend shown by the size of lower third premolar, even if not directly comparable with BM, is opposite (ca -30% at the transition P. figaro-P. sardus). Compared to P. cf. calpensis, a continental species of similar age, BM of P. figaro is ca +25%. The comparison with the insular endemic P. apricenicus evidenced differences in BM range and timespan required to attain it, due to the different size and palaeogeographical situation of the islands. Insular endemic Prolagus follow the small mammal pattern of Island Rule. Mein’s (1983) biphasic model seems applicable to the evolution of P. figaro. A tachytelic phase followed by a bradytelic one seems to characterize also the appearance of P. sardus, at least for dental traits, a process probably triggered by important variations of abiotic and biotic traits of the environment, as indicated by the turnover that marks the onset of the Dragonara subcomplex. The prediction of life history traits and other biological attributes of Sardinian Prolagus using BM should be considered with caution due to the complexity of ecological selective regimes of Sardinia.

Introduction In spite of the abundance, diversity and ubi- quity of fossil lagomorphs (leporids and ochoto- Body size is a fundamental trait in the biology nids), models for estimating the BMs of species be- and ecology of species as it shows tight correlation longing to this order were developed very recently. with several physiological, behavioral, morpholo- Quintana Cardona (2005) and Quintana et al. (2011) gical, ecological and life history attributes (Peters provided the first models for leporids. Subsequen- 1983; Calder 1984). The best proxy for quantifying tly, expanding the database of Quintana Cardona the BS of individuals is their body mass or weight (2005) and adding measurements of extant ochoto- (Gingerich et al. 1982). Thus, predicting the BM of nids, Moncunill-Solé et al. (2015) developed general fossil species is of critical significance for knowing and specific equations for estimating the BM of la- their biology as well as for understanding and gomorphs based on a multiproxy approach (teeth, quantifying their adaptations to habitats (Palombo cranial and postcranial measurements). BM estima- 2009a). For most of mammalian taxa, the allometric tion models for lagomorphs are going to enhance relationships among BM and bone/dental measu- data about palaeocommunity structures and their rements of extant relative species allow the deve- palaeoenvironmental interpretations. lopment of regression models to estimate the ave- In view of the potential of this field, we deci- rage weight of extinct ones (Damuth & MacFadden ded to study the BM of the insular endemic ocho- 1990; see Palombo 2009a: tab. 1 for a synopsis). tonids of the Pleistocene of Sardinia (Italy): Prola- gus figaro and Prolagus sardus. Prolagus figaro is known from the latest Pliocene/earliest Pleistocene to the Received: September 27, 2015; accepted: January 12, 2016 late Early Pleistocene of Sardinia (Capo Figari/ 26 Moncunill-Solé B., Tuveri C., Arca M. & Angelone C.

Femora Tibiae Humeri Tab. 1 - Fossil material used for performing the study. N N N Species Fissure filling (coding) (coding) (coding) Prolagus figaro X3 10 6 14 (SSN/X3/fe/1−10) (SSN/X3/ti/1−6) (SSN/X3/hu/1−14) IVm 5 − − (SSN/IVm/fe/1−5) X4 13 11 − (SSN/X4/fe/1−13) (SSN/X4/ti/1−11) Prolagus sardus XIr 74 42 60 (SSN/XIr/fe/1−74) (SSN/XIr/ti/1−42) (SSN/XIr/hu/1−60) VI6 20 9 49 (SSN/VI6/fe/1−20) (SSN/VI6/ti/1−9) (SSN/VI6/hu/1−49)

Orosei 1 subcomplex of the Nesogoral FC- Oro- 2) Assess the response of fossil ochotonid sei 2 subcomplex of the Microtus (Tyrrhenicola) FC; species to insular regimes (Island Rule) (see Palom- Palombo 2009b). Prolagus sardus is reported since bo 2009a and references therein for an update of the Middle Pleistocene (Dragonara subcomplex the debate about this subject) as there are not extant of the Microtus (Tyrrhenicola) FC; Palombo 2009b) relatives living on islands. until historical epoch in Sardinia and also in Cor- 3) Provide data to increase the scarce biologi- sica (Vigne & Valladas 1996; Wilkens 2004). We cal knowledge of Sardinian Prolagus. aim to: 1) Evaluate the BM trend of Prolagus in Abbreviations an insular habitat from an evolutionary point of view, as the two species of Prolagus from Sardi- BM: body mass; BS: body size; CMd1: Capo Mannu D1; IC: interval of confidence; FC: faunal complex; FL: femur length; FTDd: distal nia are part of an anagenetic evolutionary lineage femoral transversal diameter; FTDp: proximal femoral transver- (Angelone et al. 2015). sal diameter; HAPDd: distal humeral anteroposterior diameter;

Fig. 1 - BM predictions (Y axis, in g) for Prolagus figaro and P. sardus calculated on the basis of different postcranial measurements (X axis). BM average (black line), confidence interval (dotted lines) and number of individuals are shown. A)Prolagus figaro, fissure infilling X3; B)P. figaro, fissure infilling IVm; C)P. figaro, fissure infilling X4; D)Prolagus sardus, fissure infilling XIr and E)P. sardus, fissure infilling VIb. Prolagus (Ochotonidae, Lagomorpha) from the Pleistocene of Sardinia (Italy) 27

BM Prolagus figaro BM Prolagus sardus X3 IVm X4 XIr VI6 x̄ x̄ x̄ x̄ x̄ Measurement Equation N N N N N (IC) (IC) (IC) (IC) (IC) 537.57 1 444.74 1 − − 463.09 52 452.12 19 FL logBM=−1.11+2.229logFL (448.98-477.20) (426.38−477.85) 441.19 6 484.96 4 474.51 10 624.91 55 566.28 18 FTDp logBM=0.498+2.217logFTDp (337.88-544.51) (432.23-537.69) (423.18-525.84) (605.96-643.86) (538.47-594.08) 406.21 3 289.40 2 306.13 3 440.50 58 424.84 19 FTDd logBM=0.318+2.481logFTDd (207.12-605.30) (178.91-399.90) (251.76-360.51) (421.79-459.21) (398.98-450.70) 364.47 3 − − − − 633.99 47 416.71 44 HL logBM=−1.221+2.418logHL (360.73-368.22) (605.66-662.33) (403.42-430.00) 665.15 5 − − − − 633.99 37 709.82 35 HAPDp logBM=0.916+1.769logHAPDp (593.75-736.56) (605.66-662.33) (675.19-744.15) 345.03 11 − − − − 425.75 54 441.69 44 HAPDd logBM=1.354+1.769logHAPDd (311.26-378.80) (409.64-441.87) (421.16-462.25) 470.57 11 − − − − 519.21 53 528.79 44 HTDd logBM=1.053+1.513logHTDd (393.39-547.74) (493.6-544.77) (501.34-556.24) 342.28 2 − − − − 422.18 40 445.57 9 TL logBM=−1.271+2.254logTL (296.52-388.03) (407.09-437.27) (415.82-475.32) 298.77 5 − − 545.27 6 530.28 35 578.26 8 TAPDp logBM=0.599+2.265logTAPDp (233.38-364.16) (454.86-635.69) (509.28-551.47) (530.81-625.71) 170.50 4 − − 400.62 6 536.58 32 586.70 8 TTDp logBM=0.219++2.577logTTDp (155.93-185.06) (312.36-488.68) (513.18-559.98) (558.05-615.35) 334.94 2 − − 452.19 5 540.94 38 624.73 9 TTDd logBM=0.461+2.584logTTDd (241.77-428.12) (353.44-550.93) (516.22-565.66) (551.11-698.36) Arithmetic 397.88 406.37 435.74 503.86 525.05 Mean (320.68-475.08) (289.50-523.23) (357.59-513.90) (456.89-550.83) (468.12-581.97) Weighted 402.34 423.34 453.326 520.83 512.40 Average

Tab. 2 - BM predictions (in g) for the populations of Prolagus figaroand P. sardus analyzed in this paper. Last two rows, highlighted in gray, show the arithmetic mean of BM calculated for each site, and their weighted average.

HAPDp: proximal humeral anteroposterior diameter; HL: humerus also are taxonomically homogeneous infillings. The fossil material is length; HTDd: distal humeral transversal diameter; Lp3: length of curated at the SSN. the third premolar; N: sample size; SSN: Soprintendenza dei Beni Archeologici per le Provincie di Sassari e Nuoro, sede di Nuoro; TAPDp: proximal tibia anteroposterior diameter; TL: tibia length; Methods TTDd: distal tibia transversal diameter; TTDp: proximal tibia transversal diameter; : arithmetic mean;  : weighted mean. W The skeletal maturation indicates the complete cessation of longitudinal growth and, consequently, the moment when animals achieve the final BS which is maintained until their death (Peters Material 1983). Thus, BM estimations of Prolagus species were only carried out on individuals that have already attained skeletal maturity (fused The samples of P. figaro and P. sardus come from the Monte epiphyses). Specimens with unfused or broken epiphyses were not Tuttavista karstic complex (E Sardinia; Abbazzi et al. 2004) (Tab. 1). considered. Ochotona, the extant relative of Prolagus, shows a minimal Remains of P. figaro come from infillings X3, IVm and X4, pertaining sexual dimorphism (Smith 1988; Nowak 1999). For this reason, we to the Capo Figari/Orosei 1 subcomplex of the Nesogoral FC-Orosei did not assume BM differences between sexes in our fossil sample. 2 subcomplex of the Microtus (Tyrrhenicola) FC (latest Pliocene/ear- For undertaking the BM estimations, we followed the meth- liest Pleistocene to the late Early Pleistocene; Palombo 2009b). In odology described and illustrated by Moncunill-Solé et al. (2015: fig. this context, notice that Palombo (2009b) gave a different relative 1). The following measurements were taken on postcranial remains: temporal arrangement of the aforementioned infillings (IVm-X4- 1) FL, FTDd and FTDp on femora; 2) HL, HAPDp, HAPDd and X3). Remains of P. sardus have been sampled from infillings XIr and HTDd on humeri; and 3) TL, TAPDp, TTDp and TTDd on tibi- VI6, included in the Dragonara subcomplex of the Microtus (Tyrrhe- ae. Moncunill-Solé et al. (2015) observed that burrower species of nicola) FC (Middle and Late Pleistocene; Palombo 2009b). Infillings Ochotona and leporids show significant differences in the allometric XIr and VI6 were accumulated in a quite short time and their palae- models of the humerus, but not in other skeletal elements (femur ontological contents are taxonomically homogeneous (see Angelone and tibia). Although Ochotona is the closest relative of Prolagus, we et al. 2008 for discussion). Preliminary analysis of Prolagus remains preferred to use a general regression model for humeri (i.e. models and literature data based on other taxa (Angelone et al. 2009; Palom- that include data of leporids and Ochotona) because the locomotion bo 2009b and references therein) pointed out that X3, IVm and X4 of the species of Prolagus is not well known. The equations are shown 28 Moncunill-Solé B., Tuveri C., Arca M. & Angelone C.

Fig. 2 - BM range A) and Lp3 range B) showing max, average and min values of Prolagus aff. figaro (cross, CMd1), P. figaro (diamonds, X3, IVm, X4) and P. sardus (circles, IV5, IV20, XIr, VI6), with detail of average values and trends of BM (A’) and Lp3 (B’).

in Tab. 2. Once the regression models were applied, we eliminated gest N (VI6 and XIr) (Fig. 1). The variables FTDp outliers due to their potential for skewing the distributions. We fol- and HAPDp estimate a BM far above the arithme- lowed the criterion of Tukey (1977): outliers (Y) were considered when Y < (Q1−1.5IQR) or Y > (Q3+1.5IQR) (where Q1 is the 25th tic mean (between 100-200 g greater), especially in percentile, Q3 is the 75th percentile, and IQR the interquartile range VI6 population. The other variables fall next or in- (Q3−Q1)) (Quinn & Keough 2002). For each specific measurement, side the IC of the arithmetic mean (specially FL, it was calculated an arithmetic mean () and a confidence interval TAPDp, TTDp, TTDd and HTDd). Analyzing the  (IC) [ ±((σ/√N)Zα/2)]. Based on the BM of each measurement, we   results of the other populations (X3, IVm and X4), performed an arithmetic average ( ) and a weighted average ( W)

[(X1W1+X2W2+…+XNWN)/(W1+W2+…+WN)]. we observe more heterogeneous patterns. This may In order to compare the different populations of Prolagus and be consequence of: 1) few measurements taken in analyze the BM variation, we performed ANOVA analyses and post postcranial bones (3 in IVm and 5 in X4) and 2) hoc tests (Tukey HSD) (α = 0.05) using the IBM SPSS Statistics 19 software. small N (ranging from 1 to 11 individuals in X3). However, in this latter population (X3), it is already evident a large value of BM when HAPDp measu- Results rement is used, but not in FTDp. The results of BM estimations (means, IC, N) are shown in Tab. 2 and are represented in Fig. 1. Discussion For P. figaro, we estimate a weight of 397.88 g (320.68-475.08) in fissure filling X3, of 406.37 g BM of Sardinian Prolagus: trends and (289.50-523.23) in IVm and of 435.74 g (357.59- best estimators. Based on dental morphology, a 513.90) in X4. For P. sardus the results are greater, relative temporal arrangement of the studied fissure 503.86 g (456.89-550.83) in fissure filling XIr and has been attempted. Preliminary results suggested 525.05 g (468.12-581.97) in VI6. We do not observe the relative chronological arrangement X3-IVm-X4   significant differences between and W (their dif- (from older to younger) of populations of P. figa- ference is ca 10-20 g) and the latter falls perfectly in ro (Angelone et al. 2009). In the case of popula- the IC of the former (Tab. 2). Statistically, there are tions of P. sardus , infilling XIr is older than VI6 on only significant differences (p < 0.05) between the the basis of a morphological cline (Angelone et oldest population of P. figaro (X3) and the youngest al. 2008). In view of this and the BM results, the of P. sardus (VI6). three selected populations of P. figaro show a total

When the BM estimations of each measu- weight increase of ca 10% from the oldest fissure rement are assessed, a similar pattern could be filling (X3) to the youngest (X4) (see Tab. 2, Fig. observed comparing the populations with the lar- 1 and 2A). The BM of the oldest population of P. Prolagus (Ochotonidae, Lagomorpha) from the Pleistocene of Sardinia (Italy) 29

sardus here analyzed (XIr) is ca 15% greater than the that this measurement does not only represent the youngest of P. figaro (X4). The average BMs of the BM of the species but also other biological attribu- two populations of P. sardus selected for this study tes, such as locomotion or phylogeny. Samuels & show a very slight difference (average BM of VI6 is Valkenburgh (2008) described some skeletal specia- about 4% larger). Finally, the total increase among lizations of rodents depending on their locomotion the oldest (X3) to the youngest populations (VI6) style. For example, a broad and robust distal hume- of Sardinian Prolagus is of 32% (statistically signifi- rus is indicative of fossorial or semifossorial habits. cant, p < 0.05). Thus, we can affirm that Sardinian We encourage the scientific community to perform Prolagus increased its BM (average) throughout the new studies that analyze the locomotion, biomecha- Pleistocene. nics and skeletal proportions of Prolagus species in The best BM estimator for an extinct species comparison with its extant relatives (Ochotona spp.). not only depends on the accuracy of the model This will increase the biological knowledge of Prola- (statistical values), but also on a subjective judg- gus and might help us to discard those measures that ment of the results of predictions (Reynolds 2002). are correlated with their locomotion or phylogeny According to the fissure infilling with highest sam- for predicting BM. ple (XIr) (Fig. 1d), hindlimb bones seem to be the better BM estimators for Prolagus species (as shown BM and teeth size: the case of Sardinian also in Moncunill-Solé et al. 2015), particularly FL, Prolagus. It is interesting to compare the trends of TAPDp, TTDp and TTDd. However, HTDd also BM vs Lp3 in the P. figaro − P. sardus lineage. We gives adjusted estimations. All these measurements take into consideration p3 because it is the most predicted BM that fall inside the IC of the arithme- reliable tooth position for specific identification in tic mean and, consequently, we can consider them lagomorphs. As shown in Fig. 2b, average Lp3 in- good proxies for the estimation of BM in the genus creases (ca 13%) when the oldest population of P. Prolagus. However, N must be taken into due ac- figaro (X3) is compared to P. aff. figaro from CMd1 count. For example, the BM predicted by FL (N=1) site, the “founder” of the Sardinian lineage. Lp3 of in X3 population is far above the arithmetic mean. P. figaro shows a maximum oscillation of 6% in the It is recommendable to work with the largest sample considered populations. A drastic drop of Lp3 (al- possible in order to better represent the biological most 30%) is recorded between P. figaro (X4) and variability of the species and, thus, obtain more rea- the oldest studied population of P. sardus (IV5). listic values. The measurements regarded as the best After IV5, Lp3 values of P. sardus increase slightly BM estimators are surprising for two facts. Firstly, through time (total increase of ca 11% in the stu- zeugopods (tibiae), which are involved in the loco- died populations) following an asymptotic pattern motion and lifestyle of the animal, normally predict (see also Angelone et al. 2008). When we analyzed worse the BM of mammals (Scott 1990). Secondly, the BM variation, the first thing that we observe is the lengths of long bones are also considered as less that its record is more incomplete than for teeth: accurate than diameters or perimeters (Scott 1990). BM estimations are not available for P. aff. figaro and However, in the case of lagomorphs, the models older populations of P. sardus (Fig. 2a). Moreover, that use length or zeugopodial measurements are we have to take into due account that BM values as reliable (coefficient of determination or average have been obtained after complex data treatment, absolute per cent prediction) as those that use other whereas Lp3 are raw data. Nevertheless, it is evident postcranial elements, in contrast to other mamma- that BM and Lp3 of Sardinian Prolagus follow quite lian orders (see also Moncunill-Solé et al. 2015). different trends. The differences are not so evident Taking into consideration quantitative results, among populations of P. figaro: average BM increa- HAPDp measurement overestimates in all popula- ses of ca 10%, whereas average Lp3 fluctuates of tions the BM in Prolagus and could not be considered ca 6%. Evident discrepancies can be noticed with a reliable proxy. FTDp does not show a clear pat- the appearance of P. sardus. Average Lp3 drops of tern, being far above in the case of XIr population ca 30% between youngest P. figaro (X4) and oldest (those with the largest N), but not in others (X3, P. sardus (IV5). Then, Lp3 average increases throu- IVm, X4, VI6). The BM overestimation observed gh time in P. sardus attesting to a value of 2.06 mm when HAPDp is used for prediction is indicative (VI6) which is ca 15% smaller than in X4. Lacking 30 Moncunill-Solé B., Tuveri C., Arca M. & Angelone C.

data relative to older infillings (IV5 and IV20), we If we apply Mein’s model to Sardinian Prolagus can only state that younger ones (XIr and VI6) show lineage, the first phase should have taken place du- a higher BM average (ca 15-20%) than P. figaro (X4). ring or short after the Early/Late Pliocene bounda- Hypothesizing a dramatic BM decrease between P. ry (age of the CMd1 fossil assemblage). Indeed, P. figaroand P. sardus followed by an explosive increase aff. figaro from CMd1 shows very slight morpholo- to exceed P. figaro BM values is not realistic. The gical modifications due to endemism, evidence of most parsimonious hypothesis is that BM followed its very recent arrival from mainland (Angelone et a general increase trend through the transition P. al. 2015). The Lp3 of P. aff. figaro is comparable to figaro-P. sardus and throughout the evolution of P. the values of populations of continental Italy from sardus, countertrending Lp3 drastic drop observed MN16 (absence record for MN15; Angelone et al. at the transition P. figaro − P. sardus. 2015) and is between 7-13% smaller than P. figaro. The fact that p3 dimensional trend shows evi- There is no record of the possible changes of BM dent discrepancies with BM pattern inferred throu- occurred in the 1 Ma that separate CMd1 and the ol- gh postcranial elements casts doubts about the usa- dest populations of P. figaro from Monte Tuttavista. ge of p3 as a proxy for BM estimation in insular We have not enough data to clearly recognize the endemic Prolagus. Compared to continental species tachytelic stage of Mein’s model in P. aff. figaro-P. of Prolagus, insular endemic species show a notice- figaro and to verify/quantify the dimensional chan- able enlargement of the size of p3 vs the size of ges and the time span needed to produce them. The molariform elements of the lower tooth row (see populations of P. figaro here analyzed should cover a Angelone 2005: fig. 6 for a qualitative comparison) time span of ca 0.3-0.4 Ma (inferred from Palombo probably due to a reassessment in jaw mechanics. 2009b: fig 2). In this time span, slight weight fluctua- At any rate, the reliability of teeth as BM proxies tions have been observed, which may correspond to has been questioned also in studies that took into the bradytelic phase of Mein’s model. consideration a continental species of Prolagus as The appearance of P. sardus (closely related well as a wider selection of fossil lagomorphs case to P. figaro and not an immigrant from mainland; studies (Moncunill-Solé et al. 2015). They prefer Angelone et al. 2015) occurred during the transition models based on postcranial bones, as directly rela- from the Orosei 2 subcomplex to the Dragonara ted to weight bearing. subcomplex (ca 0.8-0.7 Ma; inferred from Palombo 2009b: fig 2). This transition is characterized by the Timing and patterns of BM variations highest species turnover recorded in the Quaterna- in Sardinian Prolagus. Mein (1983) illustrated a ry of Sardinia (Palombo 2009b). Leaving aside the biphasic pattern of evolution on islands consisting reason of this dramatic change (see next section), in a first tachytelic step in which the immigrant spe- it seems to have triggered a new biphasic evolutio- cies undergoes sudden morpho-dimensional chan- nary phenomenon which follows Mein’s model too. ges corresponding to its entrance to insular selec- In general the phyletic lineages of Sardinian small tive regimes and a second step in which the taxon mammals underwent outright (geologically spea- undergoes a relatively long bradytelic phase. Mil- king), an abrupt and noticeable increase in dental lien (2006) further corroborated and “quantified” size (Abbazzi et al. 2004). Contrarily Prolagus, as sta- Mein’s rule. According to some authors (Sondaar ted above, experienced a drastic Lp3 decrease. The 1977; Alcover et al. 1981; Lister 1989, 1996), the absence of data from IV5 and IV20 fissure fillings tachytelic stage is a change in the “evolutionary di- does not allow us to quantify changes in BM. Teeth rection” (sensu Sondaar 1977; e.g. BS shift or low size and morphology in early populations of P. sar- gear locomotion) whereas the bradytelic one is a dus underwent an evolution comparable to Mein’s further continuation of the existing “direction of model first phase. The slight, asymptotic growth of the change” (ib., e.g. harvesting saving by increase P. sardus Lp3 and postcranial measurements (Ange- of hypsodonty, changes in dentognathic feeding ap- lone et al. 2008), that in our data covers the inter- paratus, or developing traits for searching fallback val between ca 0.6-0.4 Ma (inferred from Palombo resources). Evans et al. (2012) estimated a minimum 2009b: fig. 2), parallelizes Mein’s model second pha- of 4000 years for small mammals to become giants se. Indeed, Mein’s rule focuses on the first stages of (ca 16000 generations). colonization of the island, making reference to the Prolagus (Ochotonidae, Lagomorpha) from the Pleistocene of Sardinia (Italy) 31

biological adaptation of the species to the new se- requirements, and the arrival of Tyrrhenicola already lective regimes. However later on, changes can also occurred earlier. The arrival of the canid Cynotheri- take place consequence of the variation in the en- um (once regarded as a specialized Prolagus hunter, vironment. Abiotic changes (climatic, topographic, and recently considered a small-prey hunter, pos- among others) and variations of biotic traits (e.g. sibly also birds, without evident specialization in levels of predation, intra- and interspecific compe- Prolagus hunting; Malatesta 1970; Lyras et al. 2006, tition), both have a significant role to drive evolu- 2010) occurred at the onset of the Orosei 2 sub- tion (Alcover et al. 1981; Brockhurst et al. 2014). complex without triggering any sudden, evident Our data seem to indicate that Mein’s model can change in small mammals, least of all in Prolagus, be applied several times to a taxon during its evo- which increases its BM throughout Pleistocene. lution of an island, if significant ecological changes This fact apparently contradicts van der Geer et al. occur (e.g. climatic changes or variation in levels of (2013) who noticed that the BS increase in insular selective regimes). The study of other mammalian small mammals that occurs following colonization lineages of the Pleistocene of Sardinia or other or first appearance, ceases or is reversed after the islands may provide more case studies to support arrival of mammalian predators or competitors. this hypothesis. Probably the impact of a new predator was not so catastrophic because, contrarily to other islands, Driving factors in the evolution of Sardi- several carnivores were already present in Sardinia nian Prolagus: some hypotheses. The sea level prior to Cynotherium (i.e. Chasmaporthetes, Mustela and low stand at the Early/Late Pliocene transition al- Pannonictis; the latter also coexisted with Cynotherium lowed the migration of P. sorbinii from Italian main- for a while). In fact, due to its large area, Sardin- land towards Sardinia (Angelone & Kotsakis 2001; ia had selective regimes more similar to mainland Angelone et al. 2015). Insular selective regimes trig- than other Mediterranean islands. It could support gered the morpho-dimensional changes in the im- the presence of terrestrial predators and had not a migrant that led to P. figaro. Prolagus figaro survived strong resource limitation as small islands (Heaney until the end of the Orosei 2 subcomplex. The ap- 1978, 1984). Thus, we can not affirm that the arrival pearance of Prolagus sardus marks the onset of the of Cynotherium sardoum increased the extrinsic mor- Dragonara subcomplex, characterized by a com- tality of Sardinian pikas. plete turnover in the small mammals’ component The most important cause of the turnover at of Sardinian fauna (except for Talpa): the leporid the onset of the Dragonara subcomplex, and thus and the glirid Tyrrhenoglis did not survive the transi- the trigger of the transition P. figaro-P. sardus is likely tion; the insectivore Nesiotites and the rodents Tyr- to be climate change, in particular those related to rhenicola and Rhagamys, descendants of taxa already the mid-Pleistocene Transition. Even after Middle present in the Orosei 2 subcomplex, underwent Pleistocene, the evolution of P. sardus seem related evident modifications of teeth morphology and a to climate changes and to consequent specific mod- noticeable increase of teeth size probably coupled ifications of the environment. Preliminary data by with a BS increase (Abbazzi et al. 2004); the ochoto- Boldrini & Palombo (2010) suggested a correlation nid Prolagus underwent a decrease of Lp3 but an in- between limb length and temperature in P. sardus. crease of BM. Effects of climate on BS have been highlighted in The ancestors of the small mammal genera insular endemic fossil vertebrates of the Mediter- which survived into the Dragonara subcomplex ranean by van der Geer et al. (2013), according to were present and already showed endemic traits whom BS fluctuates over time linked to climatic since the Orosei 2 subcomplex (i.e. Tyrrhenicola) or oscillation. Also Millien & Damuth (2004) noticed at least since the Capo Figari/Orosei 1 subcomplex. the influence of geographical climatic gradients and The competition among small mammal species as climatic change through time on fossil endemic in- driving factor of the turnover between the Orosei sular species. 2/Dragonara subcomplexes can be ruled out and Regarding the extinction of Sardinian Prola- the extinction of glirids and of the leporid is not gus, it probably occurred less than 2000 years ago, likely to have triggered a competition to occupy its in the Roman period, between the arrival of Rattus niche in taxa with such a wide range of ecological rattus and the present time (Vigne 1982). Authors 32 Moncunill-Solé B., Tuveri C., Arca M. & Angelone C.

do not agree on the importance of men’s influence sicolagus gelaberti from P. crusafonti in Lomolino et al. (directly by predation and indirectly by introduction 2013 and van der Geer et al. 2013). A reliable BS of alien predators, competitors, parasites, infectious estimation is available for an insular endemic fossil diseases, modification of the landscape by agricul- leporid, Nuralagus rex (Pliocene of Menorca, Spain), tural activities, among others) to the extinction of which BM has been calculated in 8 kg (Moncunill- Prolagus. Solé et al. 2015, who reconsidered the BM estimation of 12 kg by Quintana et al. 2011). Even if not Insular endemic lagomorphs and the Is- quantifiable, the size increase with respect to its land Rule. Radical morpho-dimensional adapta- supposed continental ancestor, the relatively small- tions are observed in insular endemic organisms. In sized genus Alilepus, should have been quite remark- mammals, apart from modifications in dental, crani- able. al and limbs morphology and relative proportions, For Prolagus species it is not possible to quan- it is observed a BS trend coined as Island Rule (Fos- tify exactly the relative BM increase of insular en- ter 1964; Van Valen 1973): in general small-sized demic vs their mainland ancestors. In the case of mammals considerably increase their size, whereas Sardinian Prolagus, this is due to the lack of studies large-sized mammals show an opposite trend. This of postcranial remains of P. sorbinii, whereas in the ecogeographic rule is also observed in insular en- case of Apulian Prolagus, their continental ances- demic fossil mammals. In the Neogene-Quaternary tor is not known yet (Angelone 2007; Angelone & of Mediterranean islands and palaeoislands, insular Čermák 2015; Angelone et al. 2015). Nevertheless, gigantism and dwarfism have been the subject of we can have a gauge of the BM difference between several studies and debates (from the pioneer gen- endemic insular Prolagus and continental species tak- eral studies, e.g. Vaufrey 1929; Thaler 1973; Sondaar ing into consideration the only available BM datum 1977; Azzaroli 1982; to the most recent reviews, e.g. of a continental Prolagus, i.e. the BM of P. cf. calpen- van der Geer et al. 2010; Lomolino et al. 2013 and sis from the Late Pliocene of Spain, estimated in ca references therein). 320 g (based on average of femurs and tibiae; Mon- Lagomorphs are usually considered as small cunill-Solé et al. 2015). Thus, P. figaro from Monte mammals together with rodents and insectivores. Tuttavista X3 had a noticeably larger BM (ca 25%) Although they have a larger size than the average of than an almost coeval western European mainland the small mammals, this order is far from reaching Prolagus. One among the oldest known populations the size of the great majority of large mammals (e.g. of P. apricenicus (Cava Fina F1; BM = ca 282 g; Mon- elephants, rhinos, etc.). Their medium or interme- cunill-Solé et al. in press) had a BM ca 13% smaller diate BM undertakes a key position in ecosystems than P. cf. calpensis, not because it decreased its size (Valverde 1964) and compromises their response in an insular domain, but probably because the con- (adaptation) to island environments (Island Rule). tinental ancestor of Apulian Prolagus was a pre-Mes- Actually in extant endemic insular leporids the BM sinian, medium-sized species (see Angelone 2007; trend reported in literature is variable, but mostly di- Angelone & Čermák 2015). Later populations of rected towards a reduction of the size (Foster 1963, P. apricenicus, weighing ca 601 g (Moncunill-Solé et 1964; Lawlor 1982; Palacios & Fernández 1992; al. 2015), almost doubled the BM of P. cf. calpensis. Tomida & Otsuka 1993). In the case of ochotonids, The known average BM range of both Sar- no extant species are present on islands and their dinian species of Prolagus (397.88-525.05 g; a BM of trend is unknown. 800 g for Mesolithic P. sardus inferred by Sondaar & When we deal with insular endemic fossil van der Geer (2000) on a qualitative basis has to be lagomorphs, it is not easy to determine their actual verified) is smaller than the populations ofP. aprice- BM and its relative variation compared to the con- nicus, whose BM range is ca 280-600 g (Moncunill- tinental ancestor. This is consequence of two facts: Solé et al. 2015, in press). The other Apulian spe- 1) mainly most remains consist in teeth, whose size, cies, P. imperialis, is traditionally considered gigantic, at least in lagomorphs, does not directly reflect BM because it has the largest p3. However, our analyses (see above and Moncunill-Solé et al. 2015); and 2) and results make clear that dental remains do not di- the supposed ancestor is often unknown or wrongly rectly reflect actual BM, and sometimes even coun- identified (e.g. P. sardus from P. michauxi and Gymne- tertrend postcranial-based results. Pending a study Prolagus (Ochotonidae, Lagomorpha) from the Pleistocene of Sardinia (Italy) 33

of postcranial remains of P. imperialis, we refrain to Taking in consideration of the aforemen- make inferences about its BM. tioned, extant and extinct relative species that Millien (2011) argued that in smaller islands dwell in habitats with low levels of extrinsic mortal- the evolutionary rate is higher. This possibly ex- ity show a slower life history than expected from plains the explosive BM increase of P. apricenicus, their BS. In the case of P. figaro and P. sardus the confined to a very limited, fragmented area, in con- levels of extrinsic mortality may not be as low as in trast to the P. figaro-P. sardus trend, which lived in a the case of P. apricenicus from Gargano consequence larger island. of the presence of predators. However, the ecologi- In general, the scanty available quantitative cal selective regimes of Sardinia would not be like data indicate that both fossil leporids and ochoto- the mainland’s one. For this reason, the prediction nids follow the small mammal Island Rule pattern. of life history traits and other biological attributes They underwent a BM increase which extent is using the estimated BM should be considered with highly variable, though. This trend is not in line with caution. Probably, we would underestimate the val- the variable response observed in extant insular en- ues of these traits. The absence of histological data demic leporids (see above). of extant and extinct ochotonids encourages the studies focused on this field in order to improve the BM and life history of insular endemic biological knowledge of insular and mainland lago- Prolagus. In the last twenty years, several researches morphs. have been focused on the life history of insular fossil species, principally addressed to dwarfism (Bro- mage et al. 2002; Raia et al. 2003; Raia & Meiri 2006; Conclusions Köhler 2010; Kubo et al. 2011; Marín-Moratalla et al. 2011; Jordana et al. 2012, 2013; van der Geer et al. BMs were estimated for P. figaro from X3 2014) although, newly, investigations regarding gi- (397.88 g), IVm (406.37 g) and X4 (435.74 g); and gantism have been performed (Moncunill-Solé et al. for P. sardus from XIr (503.86) and VI6 (525.05 in press; Orlandi-Oliveras et al. in press). BM scales g). These results allow us to state a significant in- with several traits of the life history of species such crease of BM of the species Prolagus from Sardinia as life span, fecundity, age at maturity, among others throughout the Pleistocene. The best measure- (Blueweiss et al. 1978; Peters 1983; Calder 1984). ments for determining the BM of Prolagus are FL, For this reason, at first sight, we could think in pre- TAPDp, TTDp, TTDd and HTDd. In contrast, dicting some of these life history traits for P. figaro HAPDp and FTDp appear to be unreliable prox- and P. sardus using the BM results of our analysis. ies. The BM increase opposes to the pattern of the However, Moncunill-Solé et al. (in press) analyzed Lp3, which shows a drastic drop at the transition the histology and BM of one of the Apulian insular between P. figaro and P. sardus. This is due to the endemic species of Prolagus (P. apricenicus) and sug- fact that teeth are not weight-bearing elements, and gested that it had a slower life history than expected thus, their use as BM proxies is not recommended. from its BS. Histologically, the longevity is estimat- However, when the teeth dimensions are taken into ed of at least 7 years for P. apricenicus from F1 fissure account, the biphasic Mein’s model (tachytelic and filling contrasting with the 4.5 years expected from bradytelic stages) may be observed twice. This can- its BS (around 300 g). The selective regimes of in- not be confirmed with BM estimations due to the sular habitats (low levels of extrinsic mortality and absence of postcranial elements in some key sites as resource limitation) are the most probable triggers CMd1 (P. aff. figaro) and IV5 (oldest known popula- of this shift (Palkovacs 2003). This is also observed tion of P. sardus). The entrance of Cynotherium and in extant ochotonids (Ochotona spp.) that dwell in the presence of other species of carnivores during rocky habitats which are subjected to a low aver- the Pleistocene seem to not have influence on the age yearly mortality. They show a slower life history pattern of adaption to insular ecological regimes of (later age at maturity and longer longevity) than the Prolagus. species of Ochotona that live in meadows, although Currently, the absence of ochotonids on is- both groups do not have steep differences in BM lands does not allow us to know the adaptions of this (Smith 1988). group to insular ecological regimes (Island Rule). In 34 Moncunill-Solé B., Tuveri C., Arca M. & Angelone C.

the fossil record, the two species of Prolagus studied (Western Sardinia, Italy). Boll. Soc. Paleont. It., 40: 127-32. in our research and one of the two endemic insular Angelone C., Tuveri C. & Arca M. (2009) - Biocronologia del Apulian species (P. apricenicus) suggest a gigantism Plio-Pleistocene sardo: il contributo degli ochotonidi (Lagomorpha, Mammalia). Abstr. “IX Giornate di Paleon- pattern for ochotonids. However, this latter species tologia”, 5, Apricena. shows a more explosive increase of BM perhaps as Angelone C., Tuveri C., Arca C., López Martínez N. & Kot- a result that it dwelled in a smaller, fragmented area. sakis T. (2008) - Evolution of Prolagus sardus (Ochotoni- It is observed in extant and extinct species that the dae, Lagomorpha) in the Quaternary of Sardinia island environments with a lower extrinsic mortality can (Italy). Quat. Int., 182: 109-115. promote a lower life history (e.g. greater longevity Azzaroli A. (1982) - Insularity and its effects on terrestrial vertebrates: evolutionary and biogeographic aspects. In: than that expected for its BM). Thus, the estima- Montanaro Gallitelli E. (Ed.) - Palaeontology, Essential tions of life history traits taking into account the of Historical Geology: 193-213. Edizioni S.T.E.M. Muc- BM results of our research should be considered chi, Modena. with caution. Blueweiss L., Fox H., Hudzma V., Nakashima D., Peters R. & Sams S. (1978) - Relationships between body size and Acknowledgements. We are grateful to the firms that perform some life history parameters. Oecol., 37: 257-272. quarrying activities at Monte Tuttavista for their kind collaboration; Boldrini R. & Palombo M.R. (2010) - Did temperature regu- to M. Asole, P. Catte, A. Fancello, G. Mercuriu, G. Puligheddu, A. late limb length in the Sardinian endemic ochotonid Pro- Useli for the careful work of preparation of the analyzed fossils; to lagus sardus? Abstr. “Convegno in memoria di Alberto M.A. Fadda and the Superintendents F. Lo Schiavo and F. Nicosia Malatesta (1915-2007), geologo e paleontologo”: 12-13, of the Soprintendenza Archeologica della Sardegna who allowed the Roma. study of the material analyzed in this paper; to T. Kotsakis, M.R. Bromage T.G., Dirks W., Erdjument-Bromage H., Huck M., Palombo and to the reviewers S. Čermák, X. Jordana, and J. Quin- Kulmer O., Öner R., Sandrock O. & Schrenk F. (2002) tana, and to the editor L. Rook for their useful remarks. This research - A life history and climate change solution to the evolu- was supported by the Spanish Ministry of Education, Culture and Sport (AP2010−2393, B.M-S.). tion and extinction of insular dwarfs: acypriotic experi- ence. In: Waldren W.H. & Ensenyat J.A. (Eds) - World Islands in Prehistory. International Insular Investiga- References tions, V Deia International Conference of Prehistory: 420-427. Archaeopress, Oxford. Brockhurst M.A., Chapman T., King K.C., Mank J.E., Pater- Abbazzi L., Angelone C., Arca M., Barisone G., Bedetti C., son S. & Hurst D.H.H. (2014) - Running with the Red Delfino M., Kotsakis T., Marcolini F., Palombo M.R., Queen: the role of the biotic conflicts in evolution.Proc. Pavia M., Piras P., Rook L., Torre D., Tuveri N., Valli A. Roy. Soc. B, 281: 20141382. & Wilkens B. (2004) - Plio-Pleistocene fossil vertebrates Calder W.A. III (1984) - Size, function, and life history. Dover of Monte Tuttavista (Orosei, E. Sardinia, Italy): an over- Publications Inc., New York, 431 pp. view. Riv. It. Paleont. Strat., 110: 681-706. Damuth J. & MacFadden B.J. (1990) - Body size in mammalian Alcover J.A., Moyà-Solà S. & Pons-Moyà J. (1981) - Les qui- paleobiology. Estimations and biological implications. meres del passat: els vertebrats fòssils del Plio-Quater- Cambridge University Press, Cambridge, 397 pp. nari de les Balears i Pitiüses. Editorial Moll, Ciutat de Evans A.R., Jones D., Boyer A.G., Brown J.H., Costa D.P., Er- Mallorca, 260 pp. nest S.K.M., Fitzgerald E.M.G., Fortelius M., Gittleman Angelone C. (2005) - Evolutionary trends in dental morphol- J.L., Hamilton M.J., Harding L.E., Lintulaakso K., Lyons ogy of the genus Prolagus (Ochotonidae, Lagomorpha) S.K., Okie J.G., Saarinen J.J., Sibly R.M., Smith F.A., Ste- in the Mediterranean Islands. In: Alcover J.A.& Bover phens P.R., Theodor J.M. & Uhen M.D. (2012) - The P. (Eds) - Proceedings of the International Symposium maximum rate of mammal evolution. Proc. Natl. Acad. Insular Vertebrate Evolution: the Paleontological Ap- Sciences USA, 109: 4187-4190. proach. Mon. Soc. Hist. Nat. Balears: 17-26, Mallorca. Foster J.B. (1963) - The evolution of native land mammals of Angelone C. (2007) - Messinian Prolagus (Ochotonidae, Lago- the Queen Charlotte Islands and the problem of insu- morpha, Mammalia) of Italy. Geobios, 40: 407-421. larity. PhD Thesis, University of Columbia, Vancouver. Angelone C. & Čermak S. (2015) - Two new species of Prola- Foster J.B. (1964) - Evolution of mammals on islands. Nature, gus (Lagomorpha, Mammalia) from the Late Miocene of 202: 234-235. Hungary: taxonomy, biochronology and palaeobiogeog- Gingerich P.D., Smith B.H. & Rosenberg K. (1982) - Allome- raphy. Paläontol. Z., 89(4): 1023-1038. tric scaling in the dentition of primates and prediction Angelone C., Čermak S. & Kotsakis T. (2015) - The most an- of body weight from tooth size in fossils. Am. J. Phys. cient lagomorphs of Sardinia: an overview. Geobios, 48: Anthropol., 58: 81-100. 287-296. Heaney L.R. (1978) - Island area and body size of insular Angelone C. & Kotsakis T. (2001) - Rhagapodemus azzarolii n. mammals: evidence from the tri-colored squirrel (Cal- sp. (Muridae, Rodentia) from the Pliocene of Mandriola losciurus prevosti) of Southeast Asia. Evolution, 32: 29-44. Prolagus (Ochotonidae, Lagomorpha) from the Pleistocene of Sardinia (Italy) 35

Heaney L.R. (1984) - Mammalian species richness on islands incisor of Microtia, a genus of endemic murine rodents on the Sunda Shelf, Southeast Asia. Oecol., 61: 11-17. from the late Neogene of Gargano, southern Italy. Paleo- Jordana X., Marín-Moratalla N., De Miguel D., Kaiser T.M. biology, 27: 379-391. & Köhler M. (2012) - Evidence of correlated evolution Moncunill-Solé B., Orlandi-Oliveras G., Jordana X., Rook L. of hypsodonty and exceptional longevity in endemic in- & Köhler M. (in press) - First approach of the life his- sular mammals. Proc. R. Soc. Biol. Sci., Ser. B, 279: 3339- tory of Prolagus apricenicus (Ochotonidae, Lagomorpha) 3346. from Terre Rosse sites (Gargano, Italy) using body mass Jordana X., Marín-Moratalla N., Moncunill-Solé B., Bover P., estimation and paleohistological analysis. C. R. Palevol., Alcover J.A. & Köhler M. (2013) - First fossil evidence http://dx.doi.org/10.1016/j.crpv.2015.04.004. for the advance of replacement teeth coupled with life Moncunill-Solé B., Quintana J., Jordana X., Engelbrektsson P. history evolution along an anagenetic mammalian lin- & Köhler M. (2015) - The weight of fossil leporids and eage. PLoS ONE, 8: e70743. ochotonids: body mass estimation models for the order Köhler M. (2010) - Fast or slow? The evolution of life history Lagomorpha. J. Zool., 295: 269-278. traits associated with insular dwarfing. In: Pérez-Mellado Nowak R.M. (1999) - Walker’s Mammals of the World Volume V. & Ramon C. (Eds) - Islands and Evolution: 261-280. II. Johns Hopkins University Press, Baltimore, 2015 pp. Institut Menorquí d’Estudis, Maó, Menorca. Orlandi-Oliveras G., Jordana X., Moncunill-Solé B.& Köhler Kubo M.O., Fujita M., Matsu’ura S., Kondo M. & Suwa G. M. (in press) - Bone histology of the giant fossil dor- (2011) - Mortality profiles of Late Pleistocene deer re- mouse Hypnomys onicensis (Gliridae, Rodentia) from Bale- mains of Okinawa Island: evidence from the Hananda- aric Islands. C. R. Palevol. http://dx.doi.org/10.1016/j. Gama cave and Yamashita-cho cave I sites. Anthropol. crpv.2015.05.001. Sci., 119: 183-201. Palacios F. & Fernández J. (1992) - A new subspecies of hare Lawlor T.E. (1982) - The evolution of body size in mammals: from Majorca (Balearic Islands). Mammalia, 56: 71-85. evidence from insular populations in Mexico. Am. Nat., Palombo M.R. (2009a) - Body size structure of Pleistocene 119: 54-72. mammalian communities: what support is there for the Lister A.M. (1989) - Rapid dwarfing of red deer on Jersey in “island rule”? Integr. Zool., 4: 341-356. the Last Interglacial. Nature, 342: 539-542. Palombo M.R. (2009b) - Biochronology, palaeobiogeography Lister A.M. (1996) - Dwarfing in island elephants and deer: and faunal turnover in western Mediterranean Cenozoic processes in relation to time of isolation. Symp. Zool. Soc. mammals. Integr. Zool., 4: 367-386. London, 69: 277-292. Peters R.H. (1983) - The ecological implications of body size. Lomolino M.V., van der Geer A.A.E., Lyras G.A., Palombo Cambridge University Press, Cambridge, 329 pp. M.R., Sax D.F. & Rozzi R. (2013) - Of mice and mam- Quinn G.P. & Keough M.J. (2002) - Experimental design and moths: generality and antiquity of the island rule. J. Bio- data analysis for biologists. Cambridge University Press, geogr., 40: 1427-1439. New York, 537 pp. Lyras G, Van der Geer A.A.E, Dermitzakis M.D. & De Vos J. Quintana Cardona J. (2005) - Estudio morfológico y funcional (2006) - Cynotherium sardous, an insular canid (Mammalia: de Nuralagus rex (Mammalia, Lagomopha, Leporidae). Carnivora) from the Pleistocene of Sardinia, and its ori- Unpublished PhD thesis, Universitat Autònoma de Bar- gin. J. Vert. Paleont., 26: 735-745. celona. Lyras G, Van der Geer A.A.E. & Rook L. (2010) - Body size Quintana J., Köhler M. & Moyà-Solà S. (2011) - Nuralagus rex, of insular carnivores: evidence from the fossil record. J. gen. et sp. nov., an endemic insular giant rabbit from the Biogeogr., 37: 1007-1021. Neogene of Minorca (Balearic Islands, Spain). J. Vert. Marín-Moratalla N., Jordana X., García-Martínez R. & Köhler Paleontol., 31: 231-240. M. (2011) - Tracing the evolution of fitness components Raia P., Barbera C. & Conte M. (2003) - The fast life of a in fossil bovids under different selective regimes. C. R. dwarfed giant. Evol. Ecol., 17: 293-312. Palevol, 10: 469-478. Raia P. & Meiri S. (2006) - The island rule in large mammals: Masini F., Petruso D., Bonfiglio L. & Mangano G. (2008) - paleontology meets ecology. Evolution, 60: 1731-1742. Origination and extinction patterns of mammals in Reynolds P.S. (2002) - How big is a giant? The importance of three central Western Mediterranean island form the method in estimating body size of extinct mammals. J. Late Miocene to Quaternary. Quat. Int., 182: 63-79. Mammal., 83: 321-332. Mein P. (1983) - Particularités de l’évolution insulaire chez les Samuels J.X. & Valkenburgh B.V. (2008) - Skeletal indicators petits Mammifères. Coll. Int. C.N.R.S., 330: 189-193. of locomotor adaptations in living and extinct rodents. Millien V. (2006) - Morphological evolution is accelerated J. Morphol., 269: 1387-1411. among island mammals. PLoS Biol., 4(10): e321. Scott K.M. (1990) - Postcranial dimensions of ungulates as Millien V. (2011) - Mammals evolve faster in smaller areas. predictors of body mass. In: Damuth J. & MacFadden Evolution, 65(7): 1935-1944. B.J. (Eds) - Body size in mammalian paleobiology: esti- Millien V. & Damuth J. (2004) - Climate change and size evolu- mation and biological implications: 301-305. Cambridge tion in an island rodent species: new perspectives on the University Press, Cambridge. Island Rule. Evolution, 58(6): 1353-1360. Smith A.T. (1988) - Patterns of pika (genus Ochotona) life his- Millien V. & Jaeger J-J. (2001) - Size evolution of the lower tory variation. In: Boyce M.S. (Ed.) - Evolution of Life 36 Moncunill-Solé B., Tuveri C., Arca M. & Angelone C.

Histories: Theory and Patterns from Mammals: 233- 256. Yale University Press, New Haven. Sondaar P.Y. (1977) - Insularity and its effect on mammal evolution. In: Hecht M.K., Goody P.C. & Hecht B.M. (Eds) - Major patterns in vertebrate evolution: 671-707. Ple- num Publishing Corporation, New York. Sondaar P.Y. & van der Geer A.A.E. (2000) - Mesolithic environment and animal exploitation on Cyprus and Sardin- ia/Corsica. In: Mashkour M., Choyke A.M., Buitenhuis H. & Poplin F. (Eds) - Archaeozoology of the Near East IVA: 67-73. ARC Publications 32, Groningen. Thaler L. (1973) - Nanisme et gigantisme insulaires. La Re- cherche, 37: 741-750. Tomida Y. & Otsuka H. (1993) - First Discovery of Fossil Amami Rabbit (Pentalagus furnessi) from Tokunoshima, Southwestern Japan. Bull. Nat. Sci. Mus. Tokyo, Ser. C, 19: 73-79. Tukey J.W. (1977) - Exploratory Data Analysis. Addison-Wes- ley, Boston, 688 pp. Valverde J.A. (1964) - Estructura de una comunidad de verteb- rados terrestres. Mon. Est. Biol. Doñana, 1: 1-129. van der Geer A.A.E., de Vos J., Dermitzakis M. & Lyras G. (2010) - Evolution of Island Mammals: Adaptation and Extinction of Placental Mammals on Islands. Wiley- Blackwell, Oxford, 496 pp. van der Geer A.A.E., Lyras G.A., Lomolino M.V., Palombo M.R. & Sax D.F. (2013) - Body size evolution of palaeo- insular mammals: temporal variations and interspecific interactions. J. Biogeogr., 40: 1440-1450. van der Geer A.A.E., Lyras G., MacPhee R.D.E., Lomolino M. & Drinia H. (2014) - Mortality in a predator-free insular environment: the dwarf deer of Crete. Am. Mus. Novit., 3807: 1-26. Van Valen L. (1973) - Pattern and the balance of nature. Evol. Theory, 1: 31-49. Vaufrey R. (1929) - Les éléphants nains des îles mèditerra- néennes et la question des isthmes pléistocènes. Arch. Inst. Pal. Hum., 6: 1-220. Vigne, J.-D. (1982) - Zooarchaeology and the biogeographical history of the mammals of Corsica and Sardinia since the last ice age. Mammal Rev., 22: 87-96. Vigne J.D. & Valladas H. (1996) - Small mammal fossil assem- blages as indicators of environmental change in north- ern Corsica during the last 2500 Years. J. Archaeol. Sci., 23: 199-215. Wilkens B. (2004) - La fauna sarda durante l’Olocene: le co- noscenze attuali. Sard. Cors. Bal. Antiquae, 1: 181-197.